A large peptidome dataset improves HLA class I epitope prediction across most of the human population

Sarkizova, Siranush; Klaeger, Susan; Le, Phuong M.; Li, Letitia W.; Oliveira, Giacomo; Keshishian, Hasmik; Hartigan, Christina R.; Zhang, Wandi; Braun, David A.; Ligon, Keith L.; Bachireddy, Pavan; Zervantonakis, Ioannis K.; Rosenbluth, Jennifer M.; Ouspenskaia, Tamara; Law, Travis; Justesen, Sune; Stevens, Jonathan; Lane, William J.; Eisenhaure, Thomas; Zhang, Guanglan; Clauser, Karl R.; Hacohen, Nir; Carr, Steven A.; Wu, Catherine J.; Keskin, Derin B.

doi:10.1038/s41587-019-0322-9

Cited by 380 publications

(704 citation statements)

References 48 publications

Supporting

Mentioning

656

Contrasting

Unclassified

Order By: Relevance

“…These included primary normal and chronic lymphocytic leukemia (CLL) B cells, patientderived primary glioblastoma (GBM) and melanoma cell cultures, primary healthy melanocytes, as well as established colon carcinoma and melanoma cell lines. These also included B721.221 cells, the parental cell line previously used to generate 92 single HLA allele-expressing lines from which we collected mono-allelic MHC I immunopeptidome data Sarkizova et al 2019) (Figure 1b, Supplementary Figure 1a). We developed a hierarchical ORF prediction pipeline, where ORF predictions were carried out at multiple prediction nodes, consisting of each sample (leaf), tissue (clade) and across all samples combined (root) (Figure 1c, Supplementary Figure 1a, Methods).…”

Section: A Comprehensive Pipeline For Ribo-seq Based Nuorf Identificamentioning

confidence: 99%

“…Next, we searched the MHC I immunopeptidome MS/MS spectra from 92 HLA alleles expressed in B721.221 cells (Sarkizova et al 2019) against nuORFdb with stringent FDR filtering (Supplementary Figure 2), and identified 6,501 high confidence (FDR<1%) peptides from 3,261 nuORFs, across various nuORF types (Figure 1e, Supplementary Figure 3a, Methods).…”

Section: Nuorf Derived Peptides Are Presented On Mhc Imentioning

confidence: 99%

“…First, immune responses have been detected against antigens derived from retained introns, alternative open reading frames (ORFs) within coding genes and antisense transcripts (Robbins et al 1997;Van Den Eynde et al 1999;Wang et al 1998). Additionally, while the MHC I immunopeptidome mainly consists of peptides derived from homeostatic protein turnover Sarkizova et al 2019), peptides can also be sourced from alternative precursors, such as defective ribosomal products (DRiPs), and presumably "non-coding" regions of the genome (Laumont et al 2016(Laumont et al , 2018Yewdell 2011). In particular, ribosome profiling (Ribo-seq), which assays mRNA translation by capturing and sequencing ribosome-protected mRNA fragments (RPFs) (Ingolia et al 2009), has detected a plethora of translated novel unannotated open reading frames (nuORFs).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Thousands of novel unannotated proteins expand the MHC I immunopeptidome in cancer

Ouspenskaia

Law

Clauser

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Tumor epitopes -peptides that are presented on surface-bound MHC I proteins -provide targets for cancer immunotherapy and have been identified extensively in the annotated protein-coding regions of the genome. Motivated by the recent discovery of translated novel unannotated open reading frames (nuORFs) using ribosome profiling (Ribo-seq), we hypothesized that cancer-associated processes could generate nuORFs that can serve as a new source of tumor antigens that harbor somatic mutations or show tumor-specific expression. To identify cancer-specific nuORFs, we generated Ribo-seq profiles for 29 malignant and healthy samples, developed a sensitive analytic approach for hierarchical ORF prediction, and constructed a high-confidence database of translated nuORFs across tissues. Peptides from 3,555 unique translated nuORFs were presented on MHC I, based on analysis of an extensive dataset of MHC I-bound peptides detected by mass spectrometry, with >20-fold more nuORF peptides detected in the MHC I immunopeptidomes compared to whole proteomes. We further detected somatic mutations in nuORFs of cancer samples and identified nuORFs with tumor-specific translation in melanoma, chronic lymphocytic leukemia and glioblastoma. NuORFs thus expand the pool of MHC I-presented, tumorspecific peptides, targetable by immunotherapies.

show abstract

Section: A Comprehensive Pipeline For Ribo-seq Based Nuorf Identificamentioning

confidence: 99%

Section: Nuorf Derived Peptides Are Presented On Mhc Imentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Thousands of novel unannotated proteins expand the MHC I immunopeptidome in cancer

Ouspenskaia

Law

Clauser

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…We refer to the other samples, in which exact MHC I restrictions were not experimentally determined, as the MULTIALLELIC samples. We divided these into two groups: MUTLIALLELIC-OLD, comprised of 56 experiments from eight studies published before 2018 [12][13][14][15][16][17][18][19] , and MULTIALLELIC-RECENT, comprised of 20 experiments from two studies published in 2019 10,20 .…”

Section: Ms Benchmark Construction and Approachmentioning

confidence: 99%

“…For each sample, we randomly selected 99n decoy peptides, where n is the number of hits. Equal numbers of decoy peptides of each length (8,9,10,11) were sampled.…”

Section: Ms Benchmark Construction and Approachmentioning

confidence: 99%

A model of antigen processing improves prediction of MHC I-presented peptides

O’Donnell

Rubinsteyn

Laserson

2020

Preprint

View full text Add to dashboard Cite

Large surveys of peptides naturally presented on major histocompatibility class I (MHC I) proteins have enabled improved MHC I ligand prediction by dramatically expanding the available data for many MHC I alleles. However, it is unclear to what extent antigen processing signals can also be learned from these datasets. Here, we developed a predictor of antigen processing by training neural networks to discriminate mass spec-identified MHC I ligands from unobserved peptides, where both classes of peptides are predicted to be strong MHC I binders. The resulting predictor shows qualitative consistency with established preferences for the transporter associated with antigen processing, proteasomal cleavage, and endoplasmic reticulum aminopeptidases. When we combined the antigen processing predictor with a novel pan-allele MHC I binding predictor in a logistic regression model, the combination model significantly outperformed the two components alone as well as the NetMHCpan 4.0 and MixMHCpred 2.0.2 tools at predicting mass spec-identified MHC I ligands. Our predictors are implemented in the open source MHCflurry package, version 1.6.0

show abstract

Neoantigen vaccine and neoantigen‐specific cell adoptive transfer therapy in solid tumors: Challenges and future directions

Shen¹,

Yu²,

Xu³

et al. 2022

Cancer Innovation

View full text Add to dashboard Cite

The phenomenon of tumor hierarchy and genetic instability can be explained by the “two‐hits theory” and results in the occurrence of many somatic mutations. The expression of nonsynonymous mutations results in the production of mutant proteins from tumor cells, namely tumor‐specific antigens called neoantigens. Because neoantigens do not exist in healthy cells, they have the potential to stimulate antitumor immune responses by CD4+ and CD8+ T‐cell activation without jeopardizing normal tissues. Immunotherapy has reshaped the cancer treatment paradigm in recent decades with the introduction of immune‐checkpoint blockade therapy and transgenic T‐cell receptor/chimeric antigen receptor T cells. However, these strategies performed poorly in solid tumors because of the obstacles of the immunosuppressive microenvironment caused by regulatory T cells and other suppressor cells. Therefore, other immunotherapeutic strategies are under development, such as personalized vaccines, to trigger de novo T‐cell responses against neoantigens and lead to the amplification of tumor‐specific T‐cell subclones. Neoantigen epitope prediction algorithms have enabled the detection of neoantigens and the creation of tailored neoantigen vaccines as a result of the fast development of next‐generation sequencing and cancer bioinformatics. Here we provide an overview of the current neoantigen cancer vaccines and adoptive T‐cell transfer therapy with neoantigen‐specific lymphocytes. We also discuss the challenges in developing neoantigen‐targeted immunotherapeutic strategies for cancer.

show abstract

A large peptidome dataset improves HLA class I epitope prediction across most of the human population

Cited by 380 publications

References 48 publications

Thousands of novel unannotated proteins expand the MHC I immunopeptidome in cancer

Thousands of novel unannotated proteins expand the MHC I immunopeptidome in cancer

A model of antigen processing improves prediction of MHC I-presented peptides

Neoantigen vaccine and neoantigen‐specific cell adoptive transfer therapy in solid tumors: Challenges and future directions

Contact Info

Product

Resources

About